A Model for Flood Basalt Vulcanism

The question of whether basaltic rocks in continental flood basalt provinces are primary magmas or whether they are descended in general from picritic parent magmas is reviewed. It is suggested that the latter is more likely to be correct on the evidence of phase relations and the relative rareness of mantle materials with appropriate Fe/Mg ratios. Major element variations in the residual liquids of fractional and equilibrium crystallization of basaltic magmas are modelled for a variety of crystallizing assemblages. It is concluded that crystallization of olivine, clinopyroxene, and plagioclase has a marked effect on buffering chemical change in many important elements. It is this effect which accounts for the apparent uniformity of large volumes of flood basalts, not, as has sometimes been supposed, a series of implausible coincidences in the amount of material fractionated from each magma batch. It is further argued that much of the variation seen in basalts may be imposed by polybaric fractionation operating throughout crustal depths, that is at pressures up to at least 12 kb. Parental picritic magmas rising from the mantle reach the surface in exceptional areas of crustal thinning. More usually, however, it is suggested that they intrude the base of the crust as a series of sills which differentiate into upper gabbroic and lower ultramafic portions. Much of the 'low pressure' fractionation of basaltic magmas may take place in this deep crustal sill complex and evolved liquids are transmitted to the surface as their density becomes sufficiently low. This implies that in areas of flood vulcanism a potentially large new contribution to the crust is made by underplating, the volumes of concealed cumulates being at least as large as the amount of erupted surface lava. I N T R O D U C T I O N T H E comparatively uniform and voluminous continental flood basalt sequences of the great Mesozoic-Tertiary provinces (Columbia River, North Atlantic, Deccan, Parana, Karroo, Antarctica, and Siberian Platform) pose a substantial problem with regard to the nature of their parental magmas. Are these essentially basaltic, so that much of the surface lava represents primary or near-primary magma, or are the parental magmas more magnesian (picritic), so that the surface lavas are largely derivative and non-primary? A closely similar problem is raised for the largely alkaline flood basalts of Ethiopia and Kenya as well as for midocean ridge basalts. The general topic has been the subject of considerable debate for both the oceanic and continental environments (O'Hara, 1965, 1968a; Wilkinson & Binns, 1977; Hart & Davis, 1978). This paper considers evidence from continental provinces, particularly from southern Africa, and concludes that the picritic-parent hypothesis is strongly favoured by the evidence at present available. This conclusion has far-reaching consequences for the nature of flood basalt vulcanism in general and for crustal evolution in regions affected by such episodes. (Journal of Petrology, Vol. 21, Pan 4, pp. 629-650, 19801